A human-ACE2 knock-in mouse model for SARS-CoV-2 infection recapitulates respiratory disorders but avoids neurological disease associated with the transgenic K18-hACE2 model

Word count: 245 33 Text word count: 4955 34 35 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 3

36 Animal models have been instrumental in elucidating the pathogenesis of SARS-CoV-2 37 infection and testing COVID-19 vaccines and therapeutics. Wild-type (WT) mice are not 38 susceptible to many SARS -CoV-2 variants, therefore transgenic K18 -hACE2 mice have 39 emerged as a standard model system. However, this model is characterized by severe 40 disease, particularly associated with neuroinfection, which leads to early humane endpoint 41 euthanasia. Here, we established a n ovel knock-in (KI) mouse model by inserting the 42 original K18-hACE2 transgene into the collagen COL1A1 locus using a recombinase 43 mediated cassette exchange (RMCE) system. Once the Col1a1 -K18-hACE2 mouse colony 44 was established, animals were challenged with a B.1 SARS-CoV-2 (D614G) isolate and were 45 monitored for up to 14 days. Col1a1-K18-hACE2 mice exhibited an initial weight loss similar 46 to the K18-hACE2 transgenic model but did not develop evident neurologic clinical signs. 47 The majority of Col1a1-K18-hACE2 mice did not reach the preestablished humane endpoint, 48 showing progressive weight gain after 9 days post-infection (dpi). Importantly, despite this 49 apparent milder pathogenicity compared to the K18-hACE2 transgenic model, high levels of 50 viral RNA were detected in lungs, oropharyngeal swab, and nasal turbinate. Conversely, in 51 sharp contrast to K18-hACE2 transgenic mice, no viral replication was detected in the brains 52 of Col1a1-K18-hACE2 animals at any timepoint, explaining the reduced severity of clinical 53 signs. At 14 dpi , while infection was cleared in the lungs , increased lesions and residual 54 inflammation were detected. Overall, Col1a1-K18-hACE2 mice constitute a n ew model for 55 investigating SARS-CoV-2 pathogenesis and treatments , with potential implications for 56 studying long-term COVID-19 sequelae. 57 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 4 Importance 58 K18-hACE2 mice express high levels of the human protein ACE-2, the receptor for SARS-59 CoV-2, and therefore are infected by this virus. These animals have been crucial to 60 understand viral pathogen esis and to test COVID -19 vaccines and antiviral drugs. 61 However, K18-hACE2 rapidly die after infection with initial SARS-CoV-2 variants due to 62 a massive brain infection that does not occur in humans. Here, we used a technology 63 known as knock-in that allows for the targeted insertion of a gene into a mouse and we 64 have generated a new hACE2 -mouse. We have characterized this new animal model 65 demonstrating that the virus replicates in the respiratory tract, damaging and inflaming 66 the lungs; however, in contrast to K18 -hACE2 mice, no brain infection was observed , 67 and most animals recovered from infection. This new model could be instrumental for 68 the study of specific disease aspects such as post -COVID condition, sequelae, and 69 susceptibility to reinfection. 70 71

animal model, lung inflammation, lung damage, neuroinvasion, post covid 72 condition. 73 74 75 76 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 5

77 Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) is the etiologic agent of the 78 coronavirus disease 2019 (COVID-19). The COVID-19 pandemic fuelled an unprecedented 79 collaborative research effort to develop therapeutics and vaccines . Animal models that 80 recapitulate key clinical and pathological features of COVID-19 have played a pivotal role in 81 testing novel vaccines, antivirals and other treatments (1). 82 The spike glycoprotein of SARS-CoV-2 uses the angiotensin-converting enzyme 2 (ACE2) as 83 receptor to ent er and infect host cells (2). ACE2 is expressed on the surface of cells of 84 different organs such as lungs, intestine, kidneys, and heart . However, the spike 85 glycoprotein from ancestral SARS-CoV-2 strains does not efficiently bind to mouse ACE2 86 (mACE2), rendering wild -type (WT) mice refractory to infection . Therefore, several 87 strategies were used to develop mouse models susceptible to infection . These strategies 88 include viral adaptation (3) or the introduction of the human ACE2 (hACE2) receptor via viral 89 vectors and transgenic approaches (4–8). 90 Both K18-hACE2 transgenic mice and Golden Syrian hamster s (GSH) became reference 91 animal models for investigating SARS-CoV-2 pathogenesis in vivo . They mainly differ in 92 pathogenesis: GSH, which are naturally susceptible to SARS -CoV2 infection, recapitulate a 93 milder disease phenotype (1). The transgenic mouse model was originally developed for the 94 study of SARS-CoV, which also targets human ACE2. It was created by random insertion of 95 multiple copies of the hACE2 gene under the control of the human cytokeratin 18 gene 96 promoter (KRT18 or K18). This promoter allows for high-level expression and is specific to 97 epithelial cells, including those in the airways (9, 10). This approach allows the infection of 98 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 6 mice by SARS-CoV-2, while maintaining the activity of the mACE2 receptor. Infection of K18-99 hACE2 mice with pre-Omicron variants of SARS-CoV-2 courses with progressive weight-loss 100 and strong clinical signs by 3-5 days post -infection (dpi), a severity which requires 101 euthanasia by 5-7 dpi. The lethality in this model is dose - and SARS-CoV-2 variant- 102 dependent (10, 11) , and has been associated with neuroinvasion and extensive brain 103 infection (McCray et al., 2007; Carossino et al., 2022; Tarrés-Freixas et al., 2022; Vidal et al., 104 2022). Neuroinvasion in K18 -hACE2 transgenic mice has been linked to the aleatory gene 105 insertion and the high number of inserted gene copies , around 8 in the commercial B6.Cg-106 Tg(K18-ACE2)2Prlmn/J-K18-hACE2 mouse (https://www.jax.org/strain/034860), leading to 107 altered expression levels and tissue distribution of the receptor (15). Specifically, a higher 108 number of gene copies correlates with a worse disease prognosis and encephalitis in this 109 model, following SARS-CoV infection (16). These neurological lesions are not consistent with 110 the pathology of COVID -19 in humans, where infection of the central nervous system was 111 not observed at such level (17). Therefore, new animal models that better recapitulate the 112 human disease, without viral neuroinvasion, should be developed. 113 Knock-in (KI) models are characterized by the targeted insertion of a defined gene copy 114 number in a specific locus , therefore providing a more predictable expression of the 115 transgene. In this work, we used a recombinase-based approach to insert the original K18-116 hACE2 transgene into the collagen type I alpha chain (COL1A1) locus to generate a Col1a1-117 K18-hACE2 KI mouse. To characterize this new model, mice were challenged with a SARS-118 CoV-2 B.1 (D614G) isolate and pathogenicity was compared with the well-characterized 119 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 7 K18-hACE2 transgenic mice. After viral challenge of a Col1a1 -K18-hACE2 mice , we 120 confirmed viral replication and lesions in the lungs, but absence of viral neuroinvasion. 121 122 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 8

123 Characterization of hACE2 expression in the Col1a1-K18-hACE2 mice 124 We generated a new hACE2 KI mouse model through addition of the original K18-hACE2 125 transgene, kindly supplied by Paul M cCray (9) at the C OL1A1 locus using a recombinase 126 mediated system (Figure 1A) (18). Traditional transgenic techniques rely on the random 127 genomic insertion of an unpredictable number of transgene copies. This frequently results 128 in a highly variable level and pattern of transgene expression, due to cis acting regulatory 129 elements at the site of insertion. A single copy of a transgene, inserted into the well 130 characterised Col1A1 locus, is expected to result in a more predictable level and pattern of 131 expression of the transgene sequence. To confirm the correct expression of the transgene 132 in this KI model, we quantified hACE2 transcripts relative to the house-keeping gene GAPDH 133 by RT-qPCR in RNA samples extracted from twelve different tissues. Additionally, the same 134 analysis was performed in K18-hACE2 mice to compare both models. Expression of the 135 hACE2 transgene in K18 -hACE2 mice was similar among most tested tissues except in 136 muscle and heart, which showed lower values (Figure 1B). In contrast, hACE2 expression in 137 Col1a1-K18-hACE2 mice was more heterogenous, with the lowest expression observed in 138 the brain and liver. When comparing the relative expression of hACE2 between both 139 models, we observed higher expression in the pancreas, lymph nodes and muscle of Col1a1-140 K18-hACE2 mice. Conversely, in the kidney, brain and liver , expression was higher in the 141 K18-hACE2 mouse model. Minor differences were observed in the rest of the analyzed 142 tissues (Figure 1C). 143 Survival of Col1a1-K18-hACE2 mice following SARS-CoV-2 B.1 infection. 144 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 9 Col1a1-K18-hACE2 mice (n=27) and K18 -hACE2 mice (n= 4) were challenged intranasally 145 with a SARS-CoV-2 B.1 (D614G) isolate. Uninfected KI animals, used as a control, received 146 an intranasal dose of PBS (n= 5). Animal weight and clinical signs were monitored daily for 147 14 days after infection. The endpoints to collect samples were set at 3-, 7-, and 14 -days 148 post-infection ( dpi, n=8 Col1a1-K18-hACE2 mice per endpoint) or upon fulfilment of 149 humane endpoint criteria (Figure 2A). 150 Both animal models started losing weight a round 3 dpi, consistent with previous studies 151 with B.1 isolates in transgenic animals (Figure 2B) (13). In K18-hACE2 transgenic animals, 152 the decrease in weight was associated with a dramatic increase of clinical signs: mobility, 153 dyspnoea, and neurological effects (Supplementary Table 1). This severity fulfilled the 154 humane endpoint criteria, leading to euthanasia of all K18-hACE animals by 6 or 7dpi (Figure 155 2C). In contrast, while Col1a1-K18-hACE2 mice underwent weight loss, they did not show 156 severe clinical signs apart from a slight decrease in activity . Three Col1a1-K18-hACE2 157 animals had to be euthanized (exitus) due exclusively to a weight -loss exceeding the 20% 158 limit (humane endpoint) at 8, 9 and 10 dpi. But interestingly, all other KI animals started to 159 regain weight after 9 dpi, reaching 90% of their initial weight by 14 dpi (Figure 2B/C). These 160

indicate that B.1 SARS-CoV-2 infection in Col1a1-K18-hACE2 mice is significantly less 161 lethal compared to k18-hACE2 transgenic mice (Figure 2C). 162 To explore the extent of viral infection in both animal models, viral loads (VL, cell-free viral 163 RNA) were quantified in a large set of samples (oropharyngeal swabs, lung, brain, nasal 164 turbinate, muscle, intestine, liver, kidney, pancreas, salivary glands, lymph nodes and 165 spleen) collected from a subset of animals (n=20). Respiratory tract tissues showed 166 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 10 detectable VL in both models, while the brain showed high VL exclusively in K18 -ACE2 167 transgenic samples. In both models, VL were mostly undetected in all other tissues except 168 for low levels found in the heart and salivary glands (Supplementary Figure 1). 169 B.1 replicates in the respiratory tract of Col1a1-K18-hACE2 KI mice. 170 To fully characterize the kinetics of viral replication in the respiratory tract, s amples from 171 oropharyngeal swab, lung, and nasal turbinate were collected at 3, 7, and 14 dpi or at the 172 humane endpoint (exitus) in all Col1a1-K18-hACE2 mice . Samples from K18 -hACE2 173 transgenic mice obtained at euthanasia (6 and 7 dpi) were used as reference s. Virological 174 analysis in oropharyngeal swab, lung, and nasal turbinate revealed widespread infection in 175 both K18-hACE2 and Col1a1-K18-hACE2 mice in all mentioned tissues across two 176 independent experiments (Figure 3A). VL in the nasal turbinate were similar in both animal 177 models, with some viral RNA still detected at 14 dpi in Col1a1-K18-hACE2 mice. RNA levels 178 in oropharyngeal swabs and lung peaked at 3 dpi in Col1a1-K18-hACE2 mice and tended to 179 decay overtime up to 14 dpi . Col1a1-K18-hACE2 animals euthanized at the humane 180 endpoint at 8-10 dpi still displayed high viral RNA in lungs, which could explain their more 181 pronounced weight-loss and early fulfilment of endpoint criteria. To further confirm viral 182 clearance, titration of replication-competent virus was performed in lung samples. Similar 183 to the VL kinetics, viral titers peaked at 3 dpi, followed by a quick and significant decrease 184 (Figure 3B, 3 dpi vs 7 dpi: p = 0.00178). 185 To further characterize viral pathogenesis , formalin-fixed lung samples from both animal 186 models were analyzed by histology and immunohistochemistry (IHC) to score the presence 187 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 11 of lesions and SARS-CoV-2 antigen (nucleoprotein, NP ), respectively . Histological lesions 188 were semi-quantified as none, mild, moderate, or severe (score 0-3), and the detection of 189 viral antigen as lack, low, moderate, or high (score 0-3). In Col1a1 -K18-hACE2 KI mice, 190 histopathological analysis showed an increase in lung lesion average overtime, reaching a 191 maximum score of two at 14 dpi in most animals (Figure 3C and D). IHC detection of SARS-192 CoV-2 NP in the lung s of Col1a1-K18-hACE2 KI mice was consistent with the VL analyses, 193 showing the highest score early upon infection (3 dpi). Antigen detection decreased with 194 time showing apparent clearance by 14 dpi (Figure 3C and D). 195 To assess infection-driven inflammation, the levels of IP-10, IL-6, IFNγ, MCP-1 and MIP-1β 196 were measured by Luminex in lung samples from both animal models at the indicated 197 timepoints. While all cytokines were increased in infected animals, compared to uninfected 198 controls, IP-10 and IL -6 showed a longitudinal dynamic comparable to viral load and IHC 199 data in Col1a1-K18-hACE2 KI mice, peaking at 3 dpi and significantly decreasing by 14 dpi 200 (Figure 4). In contrast, IFNγ, MCP -1, and MIPβ peak ed at 7 dpi and d ecayed afterwards, 201 although no statistical differences were observed among the analyzed timepoints. Overall, 202 levels of inflammatory mediators at 7 dpi in Col1a1-K18-hACE2 KI mice were comparable to 203 K18-hACE2 transgenic mice at the time of euthanasia. 204 Col1a1-K18-hACE2 KI mice do not develop brain infection. 205 To investigate the discrepancy in VL observed in the brain (Supplementary Figure 1 ) and 206 the differences in survival and clinical sign s observed between each model (Figure 2C), we 207 performed a longitudinal virological and histological analysis of brain samples. VL were 208 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 12 detected at low level s in only 2 out of 8 Col1a1 -K18-hACE2 KI mice analyzed at 7 dpi and 209 importantly, no VL was found in animals reaching the humane endpoint. In contrast, all K18-210 hACE2 brain samples showed high VL at euthanasia (6-7dpi) (Figure 5A). Similar results were 211 obtained when formalin-fixed brain samples were analyzed for histopathology and IHC , 212 scored in a blinded fashion as described above. Histopathological analysis of the brain in the 213 Col1a1-K18-hACE2 model showed no lesions at any of the analyzed timepoints. In contrast, 214 brain samples from K18-hACE2 infected animals showed lesion scores of 2 and 1 (Figure 5B 215 and C). Furthermore, no detection of SARS-CoV-2 NP in the brain of Col1a1-K18-hACE2 mice 216 was observed by IHC at any timepoint or at the humane endpoint, while K18-hACE2 mice 217 showed a high density of antigen diffused throughout the tissue (score 3) at euthanasia 218 (Figure 5C). Brain lesions in the transgenic model were characterized as multifocal 219 lymphoplasmacytic meningoencephalitis and were observed alongside the detection of 220 viral antigen in the brain (Figure 5C). 221 222 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 13

223 In this study, we characterized a novel hACE2 KI murine model by evaluating the course of 224 infection with the B.1 D614G SARS-CoV-2 variant with a direct comparison to the well -225 described K18-hACE2 transgenic mouse model, in which this variant is known to be highly 226 pathogenic. The most relevant findings included the susceptibility of the KI model to SARS-227 CoV-2 infection and the reduced lethality imputed to the absence of viral replication and 228 lesions in the brain. 229 The development of this KI model was started at the beginning of the COVID-19 pandemic 230 as an alternative to the K18 -hACE2 transgenic mouse model. Even though the transgenic 231 model had already been established and used in previous SARS -CoV studies (9), the 232 availability of these animals was limited at first, as colonies needed to be restarted from 233 cryopreserved embryos. In this context, several laboratories developed new hACE2 murine 234 models for SARS-CoV-2 infection. Expression of the hACE2 transgene was achieved through 235 various methods, including directed insertion using CRISPR/Cas9 technology (19) using 236 different promoters such as Hepatocyte nuclear factor -3/forkhead homologue 4 (Hfh4) 237 promoter (20) or the mouse A CE2 promoter (6), and by infecting mice with replication -238 defective adenovirus encoding hACE2 (3). In this work , we followed a different strategy 239 wherein a single copy of the transgene was inserted in to the COL1A1 locus, but under the 240 K18 promoter as the original transgenic K18 -hACE2 model. The selection of this locus was 241 based on its known ability to provide r eliable and ubiquitous expression of inserted 242 sequences (18). Additionally, RMCE allowed the insertion of a single transgene copy which 243 should better mimic physiological expression , potentially impacting lethality through 244 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 14 reduction of the ectopic expression and modification of tissue distribution of the hACE2 245 receptor. A recent study identifying a correlation between lower hACE2 copy number and 246 reduced pathology further supports our initial hypothesis (21). 247 When comparing with K18 -hACE2 transgenic mice, B.1 infection of the new Col1a1-K18-248 hACE2 KI mouse model resulted in much less severe clinical signs with better responsiveness 249 and physical appearance and no neurological sign. These observations were consistent with 250 a lack of detectable viral replication in brain of Col1a1-K18-hACE2 KI mice, as summarised 251 in Figure 6. Focusing on the respiratory tract, viral replication was observed in both models, 252 although VL and antigen detection were in general lower at 7 dpi in Col1a1 -K18-hACE2 KI 253 mice. In contrast, lung tissue lesions and inflammation were fully comparable in both 254 models at this timepoint (Figure 6). 255 To explain the different behaviour between each model, hACE2 expression was analyzed in 256 different tissues. While expression in the lungs and nasal turbinates was comparable, the 257 brain and liver showed lower expression of hACE2 in Col1a1-K18-hACE2 KI mice. This finding 258 better mimics the physiological distribution in humans, where hACE2 expression in the brain 259 is low (22). Expression of mACE2 was not analyzed , as mice are not susceptible to the B.1 260 variant through the murine receptor, and no changes in their levels or distribution were 261 expected compared to WT mice (15, 23). Among the clinical parameters analyzed in Col1a1-262 K18-hACE2 KI infected mice, we found that initial weight-loss was similar to that of the K18-263 hACE mice but was not accompanied by the increased severity of clinical signs that 264 characterizes infection in these transgenic mice (neurological signs, lack of responsiveness, 265 and poor general appearance) (13). Consistent with this observation, infection of Col1a1 -266 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 15 K18-hACE2 KI mice coursed with a partial recovery in weight by 14 dpi in most animals. 267 Importantly, the survival rate of Col1a1 -K18-hACE2 KI mice was over 70%, with the three 268 deaths exclusively due to a weight loss of more than 20%, which was one of the humane 269 endpoints in our study. 270 Among the tissues analyzed in Col1a1-K18-hACE2 KI infected mice, we found significant 271 replication of SARS CoV-2 in the respiratory tract, mainly in the lungs. Despite viral clearance 272 in Col1a1-K18-hACE2 KI mice which was confirmed by viral titration and IHC, lung lesions 273 were similar between both models by 7dpi and remained after 14 d pi in KI animals 274 recovering from the infection. The impact of viral replication in lungs, on local expression of 275 inflammatory markers, was also confirmed and found to be comparable to the inflammation 276 profile reported in K18-hACE2 transgenic mice (24, 25). Infection of Col1a1-K18-hACE2 KI 277 mice induced higher levels of IFN -g, IL -6, IP -10, MCP -1 and MIP -1b cytokines. The 278 longitudinal analysis showed early impact on IL -6 and IP-10 levels (3 dpi) and a clear trend 279 to normalization by 14 dpi, with some residual inflammation remaining . This is consistent 280 with the presence of lesions at this timepoint, and the immune cell infiltration described by 281 21 dpi in K18-hACE2 transgenic mice surviving infection (25). 282 283 The main difference between the Col1a1 -K18-hACE2 KI and the K18 -hACE2 transgenic 284 models was found in neuroinvasion. Only residual levels of viral RNA were detected in the 285 brains of Col1a1-K18-hACE2 KI mice. Furthermore, no detectable brain lesions, such as the 286 multifocal lymphoplasmacytic meningoencephalitis reported in K18-hACE2 transgenic mice 287 (14) were detected by histology . A complete absence of SARS -CoV-2 nucleoprotein in the 288 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 16 brain was evidenced by IHC. Importantly, the three animals that reached humane endpoints 289 did not present any evidence of brain viral replication but showed enhanced lung infection. 290 These observations are consistent with the lower level of hACE2 expression in the brain of 291 Col1a1-K18-hACE2 KI compared to the K18 -hACE2 transgenic animals . Therefore, the 292 targeted insertion of the hACE2 transgene seems to be a crucial factor leading to lower 293 levels of brain infection and neurological clinical signs . However, different studies have 294 shown that neuroinvasion is only partially dependent on the expression of hACE2 receptor 295 (12), and that both delivery route and viral dose can also play a role in the magnitude of the 296 neuroinvasion (22, 26). In our case, to exclusively analyze the effect of receptor expression, 297 we kept the intranasal administration route and TCID50 dose established and characterized 298 in our own and others’ studies to ensure comparability of the results (13). 299 Our study is a preliminary characterization of a new animal model and has consequently 300 several limitations. First, the expression of hACE -2 in different tissues has only been 301 analyzed at the mRNA level, with no data on immunohistochemistry to confirm the 302 presence of the protein. Moreover, we have focused on the characterization of one viral 303 variant and a single dose . The B .1 (D614G) variant was selected for its demonstrated 304 severity in K18 -hACE2 mice and its inability to use mACE2 as an entry receptor , unlike 305 B.1.351/Beta and BA1.1/Omicron variants (13, 15) . While ancestral and e arlier variants 306 (B.1.1.7/Alpha, B.1.351/Beta B.1.617.2/Delta) were highly pathogenic in K18-hACE2 mice, 307 BA.1.1/Omicron caused a milder infection with no weight loss nor neuroinvasion, therefore 308 reducing lethality in this model (13). However, more recent Omicron subvariants such as 309 BQ.1.1, BA.5 and XBB.1.5 appear to be regaining pathogenicity in K18-hACE2 mice (27–29). 310 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 17 Increased lung infection, pro-inflammatory cytokines, and lung pathology are observed with 311 these variants, with data on neuroinvasion and neurovirulence on the latter (29). Further 312 characterization of the latest variants in the KI model would provide relevant insights into 313 the full impact of the KI insertion of the transgene and confirm the suitability of this model 314 for the study of both ancestral and emerging variants. Nevertheless, for the B.1 SARS-CoV-315 2 variant it shows similarity with GSH in viral dynamics and pathology, with no 316 neuroinvasion. This new model, however, has the upside of a wider reagent availability, 317 scarce in GSH (1), and easier animal facility allowance. 318 In summary, several mouse models for the evaluation of antivirals and vaccines have been 319 developed to date , with disease phenotypes ranging from mild to severe COVID -19-like 320 condition. None of these models, however, can fully recapitulate all aspects of the disease 321 as it occurs in humans . The development and further characterization of new animal 322 models, like the Col1a1 -K18-hACE2 model, may overcome some of these limitations and 323 provide valuable tools to study certain aspects of COVID-19. Survival of Col1a1-K18-hACE2 324 KI mice after infection with a highly pathogenic variant could provide a model which better 325 mimics human disease progression, and thus could be instrumental for the study of specific 326 disease aspects such as post-COVID condition (PCC) , sequelae, long-term effects of drug 327 therapies, and susceptibility to reinfection. 328

329 Creation of the Knock-in Model 330 A new KI mouse model, B6;129S-Col1a1tm1(K18-hACE2)Irb/Irsi (Col1a1-K18-hACE2), was created 331 by inserting the original K18 -hACE2 transgene into the collagen COL1A1 locus using a 332 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 18 recombinase-mediated cassette exchange (RMCE) FLP -FRT system in KH2 cells (18). The 333 actual insertion site lies approximately 0.3 kb downstream of the 3'UTR end of COL1A1. As 334 such, the inserted transgene remains identical to the original designed one (9) and the 335 hACE2 cDNA is under the control of the K18 promoter rather than COL1A1. Specifically, the 336 plasmid PGK -ATG-Frt was digested with EcoRV , and the sequence 337 ATCAGACGTCGCTAGCGGCGCGCCGGTACTAGT was inserted to create a multi-cloning site. 338 The plasmid containing hACE2 under the control of the K18 promoter (K18-hACE2), supplied 339 by Paul McCray, was digested with HpaI and XbaI enzymes. The resulting transgene 340 fragment was isolated and purified, and then cloned into the EcoRV -NheI sites of the 341 modified PGK-ATG-Frt plasmid to generate the targeting vector. The targeting vector was 342 then co-transfected with an Flp expression construct into KH2 cells by electroporation. The 343 cells were then placed under hygromycin selection for 9 days, after which drug resistant 344 colonies were picked, expanded, and screened for the presence of the targeted transgene. 345 Once confirmed, the modified embryonic stem cells were injected into mouse blastocysts. 346 These injected mouse blastocysts were then placed into recipient foster females via embryo 347 transfer techniques. 348 Biosafety Approval and Virus Isolation 349 The execution of SARS -CoV-2 experiments was approved by the biologic biosafety 350 committee of Germans Trias i Pujol Research Institute (IGTP) and performed at the Biosafety 351 Level 3 laboratory (BSL-3) of the Comparative Medicine and Bioimage Centre (CSB-20-015-352 M8; CMCiB, Badalona, Spain). 353 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 19 The B.1 SARS-CoV-2 isolate used in this study w as isolated from nasopharyngeal swabs of 354 hospitalized patients in Spain as described elsewhere (30, 31) . Briefly, virus es were 355 propagated in Vero E6 cells (CRL-1586; ATCC, Virginia, VA, United States) for two passages 356 and recovered by supernatant collection. The sequence of the SARS-CoV-2 variant tested is 357 deposited at the GISAID Repository with accession IDs EPI_ISL_510689. EPI_ISL_510689 was 358 the first SARS-CoV-2 virus isolated in Catalonia in March 2020 and, compared to the Wuhan/ 359 Hu-1/2019 (WH1) strain, this isolate had the S protein mutations D614G, which is associated 360 with the B.1 lineage, and R682L. Viral stocks were titrated on Vero E6 cells to use equivalent 361 TCID50/mL using the Reed-Muench method and sequential 1/10 dilutions of the viral stocks 362 as described previously (31). 363 Animal Procedures and Study Design 364 All animal procedures were approved by the Committee on the Ethics of Animal 365 Experimentation of the IGTP and were authorized by the Generalitat de Catalunya (code: 366 11222). All animal experiments followed the principles of animal welfare and the 3Rs. All 367 experiments and sample processing were performed inside the BSL-3 facility. Col1a1-K18-368 hACE2 hemizygous KI mice were produced and bred at Parc Científic de Barcelona (PCB) by 369 pairing hemizygous males for Tg(K18 -ACE2)2Prlmn (or K18 -hACE2) with non -carrier 370 C57Bl6/J females. Animals used as control, B6.Cg-Tg(K18-ACE2)2Prlmn/J (or K18 -hACE2) 371 hemizygous transgenic mice (034860, Jackson Immunoresearch, West Grove, PA, United 372 States) were bred at CMCiB in the Specific Pathogen Free Area (SPF) by pairing hemizygous 373 K18-hACE2 males with non -carrier C57Bl6/J females. The genotype of the offspring 374 regarding both K18-hACE2 and Col1a1-K18-hACE2 mice was determined by qPCR at the 375 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 20 IGTP’s Genomics Platform from tail samples . Both animal models were kept in the BSL -3 376 facility during th e whole experiment including a cclimatization period. The housing 377 conditions in the BSL-3 room were maintained as follows : a temperature of 22±2ºC, 378 humidity levels between 30-70%, 20 ACH, a 12h dark/light cycle , and access to food and 379 water ad libitum. 380 A total of 36 adult mice aged 5-11 months were used in this experiment, consisting of 32 KI 381 Col1a1-K18-hACE2 mice and 4 transgenic K18-hACE2 mice. All groups were sex balanced. 382 The infections were performed in two separate experiments between January and June 383 2022. Mice were anaesthetized with isoflurane (FDG9623; Baxter, Deerfield, IL, USA) and 384 infected with a B.1 isolate (27 Col1a1-K18-hACE2 and 4 K18-hACE2 mice). The uninfected 385 control group received PBS (5 Col1a1-K18-hACE2). 386 Infection was performed using 1,000 TCID 50 of B.1 SARS -CoV-2 isolate in 50 μl o f PBS (25 387 μl/nostril), or PBS only (25 μl/nostril) for the control group. All mice fully recovered from 388 the challenge and anaesthesia procedures. Following the challenge, body weight and clinical 389 signs were monitored daily. Eight animals per group were euthanized at days 3, 7, 14 dpi or 390 upon fulfilment of human endpoint , for viral RNA quantification and histopathological 391 analyses (Supplementary Table 2). The endpoint criteria for animal welfare were 392 established based on a body weight loss superior to 20% of the initial body weight and/or 393 the display of moderate to severe clinical signs (including neurological signs), in accordance 394 with previous studies (12–14). Categories evaluated included respiration, physical 395 appearance, lack of responsiveness and neurological changes, that were scored from 0 -2 396 depending on severity . Euthanasia was performed under deep isoflurane anaesthesia by 397 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 21 whole blood extraction via cardiac puncture and was confirmed by cervical dislocation. 398 Oropharyngeal swab, lung, brain, and nasal turbinate were collected for viral RNA 399 quantification, histological and ICH analyses. For the latter techniques, tissues were fixed 400 by immersion in 10% buffered formalin . An additional set of 9 tissues was collected to 401 further characterize the KI model both for viral RNA quantification and hACE2 receptor 402 expression. These tissues included muscle, intestine, liver, kidney, lymph nodes, spleen , 403 heart, pancreas, and salivary glands. 404 hACE2 Receptor Tissue Expression 405 Several samples from both KI and transgenic models were collected and processed to 406 analyze potential differences in hACE2 receptor expression . These samples included 407 oropharyngeal swab, lung, brain, nasal turbinate, muscle, intestine, liver, kidney, pancreas, 408 salivary glands, l ymph nodes , and spleen. Tissues were processed with a TissueLyser II 409 (85,300; QIAGEN) for further VL quantification, and transgene expression was analyzed by 410 RT-qPCR analysis. 411 SARS-CoV-2 PCR Detection and Viral Load Quantification 412 Viral RNA was quantified by RT-PCR in both the standard set of samples (oropharyngeal 413 swab, lung, brain, and nasal turbinate), and the extended set (muscle, intestine, liver, 414 kidney, pancreas, salivary glands, lymph nodes and spleen ). The c ollected tissues were 415 processed, and VL determined as described by Tarrés-Freixas et al. 2022 (13). Briefly, 416 approximately 100 mg of each tissue was collected in 1.5 m L Sarstedt tubes (72,607; 417 Sarstedt, Nümbrecth, Germany) containing 500 μl of DMEM medium (11,995,065; 418 ThermoFisher Scientific) supplemented with 1% penicillin –streptomycin (10,378,016; 419 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 22 ThermoFisher Scientific, Waltham, MA, United States). A 1.5 mm Tungsten bead (69,997; 420 QIAGEN, Hilden, Germany) was added to each tube, and samples were homogenized twice 421 at 25 Hz for 30 s using a TissueLyser II (85,300; QIAGEN) before being centrifuged for 2 min 422 at 2,000 × g. Supernatants were then stored at −80°C until analysis. 423 RNA extraction was performed by using the Viral RNA/ Pathogen Nucleic Acid Isolation kit 424 (A42352, ThermoFisher Scientific), optimized for use with a KingFisher instrument 425 (5,400,610; ThermoFisher Scientific), following the manufacturer’s instructions. PCR 426 amplification was based on the 2019-Novel Coronavirus Real-Time RT-PCR Diagnostic Panel 427 guidelines and protocol developed by the American Center for Disease Control and 428 Prevention (CDC-006-00019, v.07). Briefly, a 20 μL PCR reaction was set up containing 5 μl 429 of RNA, 1.5 μ L of N2 primers and probe (2019 -nCov CDC EUA Kit, cat alogue number 430 10,006,770, Integrated DNA Technologies, Coralville, IA, USA) and 10 μl of GoTaq 1-Step RT-431 qPCR (Promega, Madison, WI, USA). Thermal cycling was performed at 50°C for 15 min for 432 reverse transcription, followed by 95°C for 2 min and then 45 cycles of 95°C for 10 s, 56°C 433 for 15 s and 72°C for 30 s in the Applied Biosystems 7,500 or QuantStudio5 Real -Time PCR 434 instruments (ThermoFisher Scientific). For absolute quantification, a standard curve was 435 built using 1/5 serial dilutions of a SARS -CoV-2 plasmid (2019 -nCoV_N_Positive Control, 436 catalog number 10006625, 200 copies/μL, Integrated DNA Technologies), which was run in 437 parallel with all PCR determinations. Viral RNA from each sample was quantified in 438 triplicate, and the mean viral RNA concentration (in copies/mL) was extrapolated from the 439 standard curve and corrected by the corresponding dilution factor. Mouse gapdh gene 440 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 23 expression was measured in duplicate for each sample using TaqMan®gene expression 441 assay (Mm99999915_g1; ThermoFisher Scientific) as amplification control. 442 Viral Titration 443 Lung tissues were evaluated for the presence of replicative virus by titration in Vero E6 cells 444 as previously described (14, 31, 32) . Briefly, after tissue homogenization, each sample 445 underwent sequential 10-fold dilutions in duplicate, transferred onto a monolayer of Vero 446 E6 cell in a 96-well plate, and incubated at 37°C and 5% CO 2. Plates were monitored daily 447 under a microscope, and at 5 dpi, wells were evaluated for the presence of cytopathic 448 effects. The amount of infectious virus was calculated by determining the TCID50 using the 449 Reed–Muench method. 450 Histopathological and Immunohistochemical Analyses 451 Tissue samples were recovered at the designated endpoint (3,7,14 dpi or exitus) and fixed 452 by immersion in 10% buffered formalin. Lung, nasal turbinate, and brain samples were 453 routinely processed for histopatholog ical examination, with haematoxylin & eosin-stained 454 slides examined under an optical microscope in a blinded fashion. A semi -quantitative 455 approach based on the amount of inflammation (none, mild, moderate, or severe) was used 456 to assess the damage caused by SARS -CoV-2 infection in mice, following a previously 457 published scoring system (14, 33). Additionally, an IHC technique was employed to detect 458 SARS-CoV-2 NP antigen in nasal turbinate, lung, and brain sections from all animals, using a 459 rabbit monoclonal antibody (40143 -R019, Sino Biological, Beijing, China) at a 1:15,000 460 dilution. The amount of viral antigen in tissues was semi -quantitatively scored in a blinded 461 fashion (low, moderate, and high amount, or lack of antigen detection) (14, 33). 462 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 24 Cytokine quantification 463 To assess the viral -driven inflammation in lung in both animal models , the levels of IP-10, 464 IL-6, IFNγ, MCP-1 and MIP-1β cytokines were analyzed by Luminex in tissue extracts. Lung 465 samples were processed as stated in the VL quantification section and stored at -80ºC until 466 analysis. In Col1a1 -K18-hACE2 mice, cytokines were analyzed at 3,7,14 dpi and those 467 euthanized by humane endpoint criteria (exitus). Uninfected Col1a1 -K18-hACE2 and K18 -468 hACE2 infected animals were used as reference groups. 469 Cytokines were measured by Luminex xMAP technology and analyzed with xPONENT 3.1 470 software (Luminex Corporation) using the MCYTOMAG -70 kit, according to the 471 manufacturers’ pr otocol with minor modifications. Briefly, after staining the desired 472 cytokines, samples underwent an overnight incubation on a rocking shaker at 4°C, using 2% 473 PFA to ensure complete inactivation of any remaining SARS-CoV-2 particles; a fixation that 474 does not alter cytokine quantification (34). Before plate acquisition , the PFA was washed 475 away and replaced with sheath fluid. 476 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 25 Statistical Analyses 477 All figures were generated using GraphPad Prism 9.0.0. Statistical analyses were performed 478 using R v4.3. Survival Rates were estimated with Kaplan-Meier curves and compared with 479 the Log-rank test. Datasets with an abundance of data below the limit of detection, like VL, 480 were analyzed using the Peto-Peto Left-censored samples test with correction for multiple 481 comparisons. Histopathological, IHC scores and viral titrations were compared using an 482 Independence Asymptotic Generalized Pearson Chi-Squared Test for ordinal data. Cytokine 483 titers were compared by a Kruskal Wallis, with pairwise comparisons conducted using 484 Conover's non-parametric test. 485 486 487 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 26

488 The authors would lik e to acknowledge Jorge Díaz, Yaiza Rosales -Salgado, Rosa María 489 Ampudia-Carrasco, Sergi Sunyé -Casas and Mireia Martínez from the CMCiB for their 490 essential help in the BSL3 facility and the K18-hACE2 mouse colony. We also thank Marisa 491 Larramona from Parc Científic de Barcelona for her unvaluable help with the knock -in 492 mouse colony. We thank the Dormeur Fondation for their financial support for the 493 acquisition of the QuantStudio-5 real-time PCR system. 494 Author Contributions 495 AP-G, FT -F, BTrinité, and JB conceived and designed the experiments. AP-G, FT-F, and 496 BTrinité performed the animal procedures. AP-G, FT-F, MP, ER-M, DR-R, DP-Z, JM-B, JS, BT 497 performed the analytical experiments. AP-G, BTrinité, JS, JV -A, VU and JB analyzed and 498 interpreted the data. SF, BTondelli established and provided the Knock-in mouse colony. SC 499 established communication between animal facilit ies and co ntributed to the veterinary 500 report verification to allow the knock-in animal shipment to the CMCiB facility. DP-Z, JM-B, 501 DR-R, NI-U provided key reagents. AP-G, BT, and JB wrote the paper. All authors contributed 502 to the article and approved the submitted version. 503 Competing Interests /Funding 504 A.P-G was supported by a predoctoral grant from Generalitat de Catalunya and Fons Social 505 Europeu (2022 FI_B 00698). This study was funded by Grifols, the Departament de Salut of 506 the Generalitat de Catalunya (grant nos. SLD016 to J.B. and SLD015 to J.C.), the Spanish 507 Health Institute Carlos III, CERCA Programme/Generalitat de Catalunya 20 21 SGR 00452, 508 and the crowdfunding initiatives #joemcorono, BonPreu/Esclat, and Correos. NI-U is 509 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 27 supported by the Spanish Ministry of Science and Innovation (grant PID2020-117145RB-I00 510 and 10.13039/501100011033, Spain), EU HORIZON -HLTH-2021CORONA-01 (grant 511 101046118, European Union) and by institutional funding of Pharma Mar, Grifols, HIPRA, 512 and Amassence. J. Blanco has received institutional funding from Grífo ls, Nesapor Europe, 513 HIPRA and MSD . Unrelated to the submitted work, J.B. and J.C. are founders and 514 shareholders of AlbaJuna Therapeutics, SL. B.C. is founder and shareholder of AlbaJuna 515 Therapeutics, SL, and AELIX Therapeutics, SL. 516 The funders had no role in study design, data collection and interpretation, or the decision 517 to submit the work for publication 518 Data availability 519 The data supporting the findings of this study are documented within the paper and are 520 available from the corresponding authors upon request. 521 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 28 522 523 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 29

524 1. Muñoz-Fontela C, Dowling WE, Funnell SGP, Gsell PS, Riveros-Balta AX, Albrecht RA, 525 Andersen H, Baric RS, Carroll MW, Cavaleri M, Qin C, Crozier I, Dallmeier K, de Waal 526 L, de Wit E, Delang L, Dohm E, Duprex WP, Falzarano D, Finch CL, Frieman MB, 527 Graham BS, Gralinski LE, Guilfoyle K, Haagmans BL, Hamilton GA, Hartman AL, 528 Herfst S, Kaptein SJF, Klimstra WB, Knezevic I, Krause PR, Kuhn JH, Le Grand R, Lewis 529 MG, Liu WC, Maisonnasse P, McElroy AK, Munster V, Oreshkova N, Rasmussen AL, 530 Rocha-Pereira J, Rockx B, Rodríguez E, Rogers TF, Salguero FJ, Schotsaert M, 531 Stittelaar KJ, Thibaut HJ, Tseng C Te, Vergara-Alert J, Beer M, Brasel T, Chan JFW, 532 García-Sastre A, Neyts J, Perlman S, Reed DS, Richt JA, Roy CJ, Segalés J, Vasan SS, 533 Henao-Restrepo AM, Barouch DH. 2020. Animal models for COVID-19. Nature 534 586:509–515. 535 2. Jackson CB, Farzan M, Chen B, Choe H. 2021. Mechanisms of SARS-CoV-2 entry into 536 cells. Nat Rev Mol Cell Biol 2021 231 23:3–20. 537 3. Hassan AO, Case JB, Winkler ES, Thackray LB, Kafai NM, Bailey AL, McCune BT, Fox 538 JM, Chen RE, Alsoussi WB, Turner JS, Schmitz AJ, Lei T, Shrihari S, Keeler SP, 539 Fremont DH, Greco S, McCray PB, Perlman S, Holtzman MJ, Ellebedy AH, Diamond 540 MS. 2020. A SARS-CoV-2 Infection Model in Mice Demonstrates Protection by 541 Neutralizing Antibodies. Cell 182:744-753.e4. 542 4. Tseng C-TK, Huang C, Newman P, Wang N, Narayanan K, Watts DM, Makino S, 543 Packard MM, Zaki SR, Chan T, Peters CJ. 2007. Severe Acute Respiratory Syndrome 544 Coronavirus Infection of Mice Transgenic for the Human Angiotensin-Converting 545 Enzyme 2 Virus Receptor. J Virol 81:1162–1173. 546 5. Jiang R Di, Liu MQ, Chen Y, Shan C, Zhou YW, Shen XR, Li Q, Zhang L, Zhu Y, Si HR, 547 Wang Q, Min J, Wang X, Zhang W, Li B, Zhang HJ, Baric RS, Zhou P, Yang X Lou, Shi 548 ZL. 2020. Pathogenesis of SARS-CoV-2 in Transgenic Mice Expressing Human 549 Angiotensin-Converting Enzyme 2. Cell 182:50-58.e8. 550 6. Bao L, Deng W, Huang B, Gao H, Liu J, Ren L, Wei Q, Yu P, Xu Y, Qi F, Qu Y, Li F, Lv Q, 551 Wang W, Xue J, Gong S, Liu M, Wang G, Wang S, Song Z, Zhao L, Liu P, Zhao L, Ye F, 552 Wang H, Zhou W, Zhu N, Zhen W, Yu H, Zhang X, Guo L, Chen L, Wang C, Wang Y, 553 Wang X, Xiao Y, Sun Q, Liu H, Zhu F, Ma C, Yan L, Yang M, Han J, Xu W, Tan W, Peng 554 X, Jin Q, Wu G, Qin C. 2020. The pathogenicity of SARS-CoV-2 in hACE2 transgenic 555 mice. Nat 2020 5837818 583:830–833. 556 7. Zhou B, Thao TTN, Hoffmann D, Taddeo A, Ebert N, Labroussaa F, Pohlmann A, King 557 J, Steiner S, Kelly JN, Portmann J, Halwe NJ, Ulrich L, Trüeb BS, Fan X, Hoffmann B, 558 Wang L, Thomann L, Lin X, Stalder H, Pozzi B, de Brot S, Jiang N, Cui D, Hossain J, 559 Wilson MM, Keller MW, Stark TJ, Barnes JR, Dijkman R, Jores J, Benarafa C, 560 Wentworth DE, Thiel V, Beer M. 2021. SARS-CoV-2 spike D614G change enhances 561 replication and transmission. Nature 592:122–127. 562 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 30 8. Winkler ES, Chen RE, Alam F, Yildiz S, Case JB, Uccellini MB, Holtzman MJ, Garcia-563 Sastre A, Schotsaert M, Diamond MS. 2022. SARS-CoV-2 Causes Lung Infection 564 without Severe Disease in Human ACE2 Knock-In Mice. J Virol 96. 565 9. McCray PB, Pewe L, Wohlford-Lenane C, Hickey M, Manzel L, Shi L, Netland J, Jia HP, 566 Halabi C, Sigmund CD, Meyerholz DK, Kirby P, Look DC, Perlman S. 2007. Lethal 567 Infection of K18- hACE2 Mice Infected with Severe Acute Respiratory Syndrome 568 Coronavirus . J Virol 81:813–821. 569 10. Zheng J, Wong LYR, Li K, Verma AK, Ortiz ME, Wohlford-Lenane C, Leidinger MR, 570 Knudson CM, Meyerholz DK, McCray PB, Perlman S. 2021. COVID-19 treatments 571 and pathogenesis including anosmia in K18-hACE2 mice. Nature 589:603–607. 572 11. Winkler ES, Bailey AL, Kafai NM, Nair S, McCune BT, Yu J, Fox JM, Chen RE, Earnest 573 JT, Keeler SP, Ritter JH, Kang LI, Dort S, Robichaud A, Head R, Holtzman MJ, 574 Diamond MS. 2020. SARS-CoV-2 infection of human ACE2-transgenic mice causes 575 severe lung inflammation and impaired function. Nat Immunol 2020 2111 21:1327–576 1335. 577 12. Carossino M, Kenney D, O’connell AK, Montanaro P, Tseng AE, Gertje HP, Grosz KA, 578 Ericsson M, Huber BR, Kurnick SA, Subramaniam S, Kirkland TA, Walker JR, Francis 579 KP, Klose AD, Paragas N, Bosmann M, Saeed M, Balasuriya UBR, Douam F, Crossland 580 NA. 2022. Fatal Neurodissemination and SARS-CoV-2 Tropism in K18-hACE2 Mice Is 581 Only Partially Dependent on hACE2 Expression. Viruses 2022:535. 582 13. Tarrés-Freixas F, Trinité B, Pons-Grífols A, Romero-Durana M, Riveira-Muñoz E, 583 Ávila-Nieto C, Pérez M, Garcia-Vidal E, Perez-Zsolt D, Muñoz-Basagoiti J, Raïch-584 Regué D, Izquierdo-Useros N, Andrés C, Antón A, Pumarola T, Blanco I, Noguera-585 Julián M, Guallar V, Lepore R, Valencia A, Urrea V, Vergara-Alert J, Clotet B, Ballana 586 E, Carrillo J, Segalés J, Blanco J. 2022. Heterogeneous Infectivity and Pathogenesis of 587 SARS-CoV-2 Variants Beta, Delta and Omicron in Transgenic K18-hACE2 and 588 Wildtype Mice. Front Microbiol 13. 589 14. Vidal E, López-Figueroa C, Rodon J, Pérez M, Brustolin M, Cantero G, Guallar V, 590 Izquierdo-Useros N, Carrillo J, Blanco J, Clotet B, Vergara-Alert J, Segalés J. 2022. 591 Chronological brain lesions after SARS-CoV-2 infection in hACE2-transgenic mice. 592 Vet Pathol 59:613–626. 593 15. Shuai H, Chan JF-W, Yuen TT-T, Yoon C, Hu J-C, Wen L, Hu B, Yang D, Wang Y, Hou Y, 594 Huang X, Chai Y, Chan CC-S, Poon VK-M, Lu L, Zhang R-Q, Chan W-M, Ip JD, Chu AW-595 H, Hu Y-F, Cai J-P, Chan K-H, Zhou J, Sridhar S, Zhang B-Z, Yuan S, Zhang AJ, Huang J-596 D, To KK-W, Yuen K-Y, Chu H. 2021. Emerging SARS-CoV-2 variants expand species 597 tropism to murines. EBioMedicine 73:103643. 598 16. Netland J, Meyerholz DK, Moore S, Cassell M, Perlman S. 2008. Severe Acute 599 Respiratory Syndrome Coronavirus Infection Causes Neuronal Death in the Absence 600 of Encephalitis in Mice Transgenic for Human ACE2. J Virol 82:7264–7275. 601 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 31 17. Meinhardt J, Streit S, Dittmayer C, Manitius R v., Radbruch H, Heppner FL. 2023. The 602 neurobiology of SARS-CoV-2 infection. Nat Rev Neurosci 2023 251 25:30–42. 603 18. Beard C, Hochedlinger K, Plath K, Wutz A, Jaenisch R. 2006. Efficient method to 604 generate single-copy transgenic mice by site-specific integration in embryonic stem 605 cells. Genesis 44:23–28. 606 19. Sun SH, Chen Q, Gu HJ, Yang G, Wang YX, Huang XY, Liu SS, Zhang NN, Li XF, Xiong R, 607 Guo Y, Deng YQ, Huang WJ, Liu Q, Liu QM, Shen YL, Zhou Y, Yang X, Zhao TY, Fan CF, 608 Zhou Y Sen, Qin CF, Wang YC. 2020. A Mouse Model of SARS-CoV-2 Infection and 609 Pathogenesis. Cell Host Microbe 28:124-133.e4. 610 20. Jiang R Di, Liu MQ, Chen Y, Shan C, Zhou YW, Shen XR, Li Q, Zhang L, Zhu Y, Si HR, 611 Wang Q, Min J, Wang X, Zhang W, Li B, Zhang HJ, Baric RS, Zhou P, Yang X Lou, Shi 612 ZL. 2020. Pathogenesis of SARS-CoV-2 in Transgenic Mice Expressing Human 613 Angiotensin-Converting Enzyme 2. Cell 182:50-58.e8. 614 21. Bradshaw CM, Georgieva T, Tankersley TN, Taylor- T, Johnson L, Uhrlaub JL, 615 Besselsen D, Janko Ž. 2023. Characterization of low copy number human 616 angiotensin- converting enzyme 2 ( hACE2 ) -transgenic mice as an improved model 617 of SARS-CoV-2 infection 2. 618 22. Dong W, Mead H, Tian L, Park J-G, Garcia JI, Jaramillo S, Barr T, Kollath DS, Coyne 619 VK, Stone NE, Jones A, Zhang J, Li A, Wang L-S, Milanes-Yearsley M, Torrelles JB, 620 Martinez-Sobrido L, Keim PS, Barker BM, Caligiuri MA, Yu J. 2022. The K18-Human 621 ACE2 Transgenic Mouse Model Recapitulates Non-severe and Severe COVID-19 in 622 Response to an Infectious Dose of the SARS-CoV-2 Virus. J Virol 96. 623 23. Gembardt F, Sterner-Kock A, Imboden H, Spalteholz M, Reibitz F, Schultheiss HP, 624 Siems WE, Walther T. 2005. Organ-specific distribution of ACE2 mRNA and 625 correlating peptidase activity in rodents. Peptides 26:1270–1277. 626 24. Oladunni FS, Park JG, Pino PA, Gonzalez O, Akhter A, Allué-Guardia A, Olmo-627 Fontánez A, Gautam S, Garcia-Vilanova A, Ye C, Chiem K, Headley C, Dwivedi V, 628 Parodi LM, Alfson KJ, Staples HM, Schami A, Garcia JI, Whigham A, Platt RN, Gazi M, 629 Martinez J, Chuba C, Earley S, Rodriguez OH, Mdaki SD, Kavelish KN, Escalona R, 630 Hallam CRA, Christie C, Patterson JL, Anderson TJC, Carrion R, Dick EJ, Hall-Ursone S, 631 Schlesinger LS, Alvarez X, Kaushal D, Giavedoni LD, Turner J, Martinez-Sobrido L, 632 Torrelles JB. 2020. Lethality of SARS-CoV-2 infection in K18 human angiotensin-633 converting enzyme 2 transgenic mice. Nat Commun 11. 634 25. Yinda CK, Port JR, Bushmaker T, Owusu IO, Purushotham JN, Avanzato VA, Fischer 635 RJ, Schulz JE, Holbrook MG, Hebner MJ, Rosenke R, Thomas T, Marzi A, Best SM, de 636 Wit E, Shaia C, van Doremalen N, Munster VJ. 2021. K18-hACE2 mice develop 637 respiratory disease resembling severe COVID-19. PLOS Pathog 17:e1009195. 638 26. Fumagalli V, Ravà M, Marotta D, Di Lucia P, Laura C, Sala E, Grillo M, Bono E, 639 Giustini L, Perucchini C, Mainetti M, Sessa A, Garcia-Manteiga JM, Donnici L, 640 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 32 Manganaro L, Delbue S, Broccoli V, De Francesco R, D’Adamo P, Kuka M, Guidotti 641 LG, Iannacone M. 2022. Administration of aerosolized SARS-CoV-2 to K18-hACE2 642 mice uncouples respiratory infection from fatal neuroinvasion. Sci Immunol 7. 643 27. Case JB, Scheaffer SM, Darling TL, Bricker TL, Adams LJ, Harastani H, Trende R, 644 Sanapala S, Fremont DH, Boon ACM, Diamond MS. 2023. Characterization of the 645 SARS-CoV-2 BA.5.5 and BQ.1.1 Omicron Variants in Mice and Hamsters. bioRxiv 646 Prepr Serv Biol 2023.04.28.538747. 647 28. Halfmann PJ, Iida S, Iwatsuki-Horimoto K, Maemura T, Kiso M, Scheaffer SM, Darling 648 TL, Joshi A, Loeber S, Singh G, Foster SL, Ying B, Case JB, Chong Z, Whitener B, 649 Moliva J, Floyd K, Ujie M, Nakajima N, Ito M, Wright R, Uraki R, Warang P, Gagne M, 650 Li R, Sakai-Tagawa Y, Liu Y, Larson D, Osorio JE, Hernandez-Ortiz JP, Henry AR, 651 Ciuoderis K, Florek KR, Patel M, Odle A, Wong LYR, Bateman AC, Wang Z, Edara VV, 652 Chong Z, Franks J, Jeevan T, Fabrizio T, DeBeauchamp J, Kercher L, Seiler P, 653 Gonzalez-Reiche AS, Sordillo EM, Chang LA, van Bakel H, Simon V, Alburquerque B, 654 Alshammary H, Amoako AA, Aslam S, Banu R, Cognigni C, Espinoza-Moraga M, 655 Farrugia K, van de Guchte A, Khalil Z, Laporte M, Mena I, Paniz-Mondolfi AE, 656 Polanco J, Rooker A, Sominsky LA, Douek DC, Sullivan NJ, Thackray LB, Ueki H, 657 Yamayoshi S, Imai M, Perlman S, Webby RJ, Seder RA, Suthar MS, García-Sastre A, 658 Schotsaert M, Suzuki T, Boon ACM, Diamond MS, Kawaoka Y. 2022. SARS-CoV-2 659 Omicron virus causes attenuated disease in mice and hamsters. Nat 2022 6037902 660 603:687–692. 661 29. Stewart R, Yan K, Ellis SA, Bishop CR, Dumenil T, Tang B, Nguyen W, Larcher T, Parry 662 R, Sng JDJ, Khromykh AA, Sullivan RKP, Lor M, Meunier FA, Rawle DJ, Suhrbier A. 663 2023. SARS-CoV-2 omicron BA.5 and XBB variants have increased neurotropic 664 potential over BA.1 in K18-hACE2 mice and human brain organoids. Front Microbiol 665 14. 666 30. Perez-Zsolt D, Muñoz-Basagoiti J, Rodon J, Elousa M, Raïch-Regué D, Risco C, Sachse 667 M, Pino M, Gumber S, Paiardini M, Chojnacki J, Erkizia I, Muñiz X, Ballana E, Riveira-668 Muñoz E, Noguera M, Paredes R, Trinité B, Tarrés-Freixas F, Blanco I, Guallar V, 669 Carrillo J, Blanco J, Telenti A, Heyn H, Segalés J, Clotet B, Martinez-Picado J, Vergara-670 Alert J, Izquierdo-Useros N. 2021. Siglec-1 on dendritic cells mediates SARS-CoV-2 671 trans-infection of target cells while on macrophages triggers proinflammatory 672 responses. bioRxiv 8:2021.05.11.443572. 673 31. Rodon J, Muñoz-Basagoiti J, Perez-Zsolt D, Noguera-Julian M, Paredes R, Mateu L, 674 Quiñones C, Perez C, Erkizia I, Blanco I, Valencia A, Guallar V, Carrillo J, Blanco J, 675 Segalés J, Clotet B, Vergara-Alert J, Izquierdo-Useros N. 2021. Identification of 676 Plitidepsin as Potent Inhibitor of SARS-CoV-2-Induced Cytopathic Effect After a Drug 677 Repurposing Screen. Front Pharmacol 12:646676. 678 32. Haagmans BL, van den Brand JMA, Raj VS, Volz A, Wohlsein P, Smits SL, Schipper D, 679 Bestebroer TM, Okba N, Fux R, Bensaid A, Solanes Foz D, Kuiken T, Baumgärtner W, 680 Segalés J, Sutter G, Osterhaus ADME. 2016. An orthopoxvirus-based vaccine 681 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 33 reduces virus excretion after MERS-CoV infection in dromedary camels. Science (80- 682 ) 351:77–81. 683 33. Brustolin M, Rodon J, Rodríguez de la Concepción ML, Ávila-Nieto C, Cantero G, 684 Pérez M, Te N, Noguera-Julián M, Guallar V, Valencia A, Roca N, Izquierdo-Useros N, 685 Blanco J, Clotet B, Bensaid A, Carrillo J, Vergara-Alert J, Segalés J. 2021. Protection 686 against reinfection with D614- or G614-SARS-CoV-2 isolates in golden Syrian 687 hamster. Emerg Microbes Infect 10:797. 688 34. Dowall SD, Graham VA, Tipton TRW, Hewson R. 2009. Multiplex cytokine profiling 689 with highly pathogenic material: use of formalin solution in luminex analysis. J 690 Immunol Methods 348:30–35. 691 692 693 694 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 34 Figure Legends 695 Figure 1. Expression of hACE2 in the Col1a1-K18-hACE2 KI model. (A) Schematic 696 representation of insertion strategy. The original K18-hACE2 transgene was inserted into 697 the collagen C OL1A1 locus using a recombinase mediated cassette exchange (RMCE) FLP -698 FRT system in KH2 cells via blastocyst injection. pPGK-ATG-frt plasmid : vector backbone 699 with Ampicillin resistant gene (Amp), transcription start site (ATG) and Flippase recognition 700 target (Frt). Hygro: Hygromicin resistance gene. pCAGGS-FlpE: expression plasmid for FLPe 701 recombinase expression. (B) Relative quantification of hACE2 receptor expression to GAPDH 702 expression in uninfected Col1a1-K18-hACE2 ((blue empty dot, n=3) and K18-hACE2 males 703 (black empty square, n=3). Delta Ct values are inversely shown to facilitate interpretation. 704 The lower the absolute number, the higher the relative expression. Solid line and bars 705 represent mean and SEM. (C) Relative comparison of hACE2 receptor expression in Col1a1-706 K18-hACE2 versus K18-hACE2 mice (n=3 each, males). Higher receptor expression in Col1a1-707 K18-hACE2 (negative values in Y axis) is marked by blue bars, and higher expression in K18-708 hACE2 (positive values in Y axis) is marked in black bars. 709 Figure 2. Experimental setting and progression of SARS -CoV-2 infection in Col1a1 -K18-710 hACE2 and K18 -hACE2 mouse models. (A) Schematic representation of the experimental 711 setting. Knock -in Col1a1 -K18-hACE2 (n=27) and transgenic K18 -hACE2 mice (n=4) were 712 intranasally challenged with a 1000 TCID50 dose of a B.1 SARS-CoV-2 isolate. A Col1a1-K18-713 hACE2 uninfected control group (n=5) was challenged with PBS. Mice were monitored for 714 weight-loss and clinical signs for 14 days post infection (dpi). Euthanasia was performed at 715 3, 7, and 14 dpi or upon fulfilment of humane endpoint criteria , for sample and tissue 716 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 35 collection (n=8 per timepoint). Infections were performed in two separate experiments 717 between January and June 2022. Created with Biorender.com. (B) Relative body weight 718 follow-up referred to day 0 . Col1a1-K18-hACE2 uninfected (blue empty dot), Col1a1 -K18-719 hACE2 infected (blue dot), K18-hACE2 infected (black square). Solid lines and bars represent 720 mean±SD. (C) Survival ( Kaplan-Meier). All K18 -hACE2 infected animals (n=4) had to be 721 euthanized due to endpoint criteria by 7dpi, and only three infected Col1a1 -K18-hACE2 at 722 8, 9, and 10 dpi. No uninfected Col1a1-K18-hACE2 had to be euthanized under this criterion. 723 Col1a1-K18-hACE2 uninfected (blue dashed line), Col1a1 -K18-hACE2 infected (blue line), 724 K18-hACE2 infected (black line). Statistical differences were identified using a Log-rank 725 (Mantel-Cox) test ( <0.0001), followed by individual comparisons (**p < 0.0 05; ** **p < 726 0.0001) 727 Figure 3. Progression of SARS -CoV-2 infection in Col1a1-K18-hACE2 KI mice. Col1a1-K18-728 hACE2 (blue circles, n=27) and K18-hACE2+ mice (black squares, n=4) were inoculated with 729 1000 TCID50 of a SARS-CoV-2 B.1 isolate (full shapes) or uninfected (empty shapes of each 730 colour, n=5). Animals were euthanized at 3dpi (n= 8), 7dpi (n= 8), 14 dpi (n=8) or upon 731 fulfilment of humane endpoint criteria. (A) SARS -CoV-2 viral RNA loads (copies/mL) of 732 oropharyngeal swab, lung, and nasal turbinate samples. Dashed line represents limit of 733 detection, established by 2SD of uninfected animals. Statistical differences were identified 734 using a Peto-Peto Left censored test (*p < 0.05, **p<0.005). (B) Viral titration of replicative 735 virus (TCID50/mL) in lung samples from B.1 infected mice at different endpoints in Vero E6 736 cells on day 5 of culture. Titers were compared using an Independence Asymptotic 737 Generalized Pearson Chi -Squared Test for ordinal data (*p < 0.05, **p<0.005 ). (C) Lung 738 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 36 histopathological scoring of broncho -interstitial pneumonia (left) and SARS-CoV-2 NP 739 immunohistochemical ( IHC) scoring (right) in both models at 3, 7, 14 dpi and endpoint . 740 Statistical differences were identified using an Independence Asymptotic Generalized 741 Pearson Chi-Squared Test for ordinal data (*p < 0.05). (D) Representative lung histology and 742 IHC pictures of both models at 3, 7, 14 dpi and endpoint. Images show low -power 743 magnification bars (200 µm). 744 Figure 4. Inflammatory response in the lung. The concentration of Inflammatory cytokines 745 in lung extracts in the knock-in and transgenic models at 3,7,14 dpi or endpoint are shown. 746 Col1a1-K18-hACE2 infected (blue full circles, n=23), uninfected (blue empty circles, n=3) and 747 K18-hACE2+ infected mice (black full squares, n=4). The bar shows Median with 748 interquartile range . Limit of detection for each cytokine is indicated by a dotted line . 749 Statistical differences were identified using a Kruskal Wallis, and a Conover’s nonparametric 750 all-pairs comparison test (*p < 0.05, **p<0.005). 751 Figure 5. Progression of SARS-CoV-2 infection in brain from the Col1a1-K18-hACE2 KI mice. 752 Col1a1-K18-hACE2 (blue circles, n=27) and K18 -hACE2+ mice (black squares, n=4) were 753 inoculated with 1000 TCID50 of a B.1 SARS-CoV-2 isolate (full shapes) or uninfected (empty 754 shapes, n=5) and followed until 14 days post infection (dpi). Samples were collected at 3, 7, 755 14 dpi and endpoint. (A) SARS-CoV-2 viral RNA loads (copies/mL) in brain extracts. Dashed 756 line represents limit of detection, established by 2SD of uninfected animals. Statistical 757 differences were identified using a Peto-Peto Left-censored samples test with correction for 758 multiple comparisons. (B) Brain histopathological scoring of multifocal lymphoplasmacytic 759 meningo-encephalitis in brain (Left) and SARS-CoV-2 NP IHC scoring (right) in brain of both 760 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint 37 models. Statistical differences were identified using an Independence Asymptotic 761 Generalized Pearson Chi -Squared Test for ordinal data (*p < 0.05; **p < 0.01). (C) 762 Representative images o f brain Histology and IHC images of both hACE2+ mouse models. 763 Images show low-power magnification (top ; bar s: 600µm) and medium -power 764 magnification (bottom; bars: 300μm). Col1a1-K18-hACE2 KI mice samples shown were 765 collected at 7 dpi, while K18-hACE2 samples were collected at humane endpoint (6-7dpi). 766 Figure 6. Spider Plot of clinical data compilation at 7 dpi from both animal models. 767 Comparative data summary of Col1a1-K18-hACE2 and K18-hACE2 mice infection by a B.1 768 SARS-CoV-2 isolate. Data at 7dpi of the KI (blue; euthanized at 7 dpi) and K18-hACE2 (black; 769 euthanized at 6-7dpi) of different parameters was compilated and median values are shown 770 in each axis of the spider plot. Data ranges are adjusted for each kind of data, and shown by 771 concentric circles, being the outside circle the highest level. Clinical status was assessed in 772 different categories: respirat ion, appearance, weight-loss, lack of responsiveness and 773 neurological signs. Histology and IHQ data were semiquantitatively scored from 0-2, 0-3 and 774 0-3 respectively as described in Materials and Methods section . Viral loads are shown as 775 Log10 copies/mL, titration as Log10 TCID50 and cytokines as pg/mL. 776 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint Fig 1. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint Fig 2. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint Fig 3. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint Fig 4. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint Fig 5. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint Fig 6. .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted June 13, 2024. ; https://doi.org/10.1101/2024.06.11.598471doi: bioRxiv preprint